Superior Petrosal Sinus: Hemodynamic Features in Normal and Cavernous Sinus Dural Arteriovenous Fistulas

BACKGROUND AND PURPOSE: Normal hemodynamic features of the superior petrosal sinus and their relationships to the SPS drainage from cavernous sinus dural arteriovenous fistulas are not well known. We investigated normal hemodynamic features of the SPS on cerebral angiography as well as the frequency and types of the SPS drainage from CSDAVFs. MATERIALS AND METHODS: We evaluated 119 patients who underwent cerebral angiography by focusing on visualization and hemodynamic status of the SPS. We also reviewed selective angiography in 25 consecutive patients with CSDAVFs; we were especially interested in the presence of drainage routes through the SPS from CSDAVFs. RESULTS: In 119 patients (238 sides), the SPS was segmentally (anterior segment, 37 sides; posterior segment, 82 sides) or totally (116 sides) demonstrated. It was demonstrated on carotid angiography in 11 sides (4.6%), receiving blood from the basal vein of Rosenthal or sphenopetrosal sinus, and on vertebral angiography in 235 sides (98.7%), receiving blood from the petrosal vein. No SPSs were demonstrated with venous drainage from the cavernous sinus. SPS drainage was found in 7 of 25 patients (28%) with CSDAVFs. CSDAVFs drained through the anterior segment of SPS into the petrosal vein without draining to the posterior segment in 3 of 7 patients (12%). CONCLUSIONS: The SPS normally works as the drainage route receiving blood from the anterior cerebellar and brain stem venous systems. The variation of hemodynamic features would be related to the relatively lower frequency and 2 different types of SPS drainage from CSDAVFs.

T he SPS is well known as a dural sinus that communicates with the cavernous sinus anteriorly and with the transverse sinus posteriorly. It runs over the trigeminal nerve in the lateral margin of the tentorium and in the superior petrosal sulcus of the temporal bone. In some textbooks, the SPS is described as a dural sinus that receives blood from the cavernous sinus and drains into the transverse sinus. 1,2 Normal hemodynamic features of the SPS have not been reported, however, and the SPS is not frequently demonstrated at venous phase on internal carotid angiography in our clinical practice.
Furthermore, cavernous sinus dural arteriovenous fistulas less frequently drain via the SPS to the transverse sinus. [3][4][5] In this study, we investigated the normal hemodynamic feature of the SPS on cerebral angiography as well as the frequency and types of SPS drainage from CSDAVFs.

MATERIALS AND METHODS
For evaluating normal hemodynamic features of the SPS, we retrospectively reviewed 119 consecutive patients who underwent cerebral angiography between May 2010 and October 2011 at our institution. Patients who had lesions affecting cerebral venous drainage were excluded from this study. Patient age ranged from 12 to 84 years (mean age, 56.6 years), and there were 56 males and 63 females. Selective digital subtraction angiography of the bilateral common/internal carotid arteries and the vertebral arteries was performed in all patients without breath-holding by using biplane angiography equipment (Infinix Celeve-i INFX-8000V; Toshiba Medical Systems, Tokyo, Japan). Two experienced neuroradiologists evaluated the images carefully by focusing on visualization of the SPS and its hemodynamic status.
We also reviewed selective angiography from 25 consecutive patients with CSDAVFs in whom angiography was performed between September 2003 and January 2011 at our institution. There were 4 men and 21 women, ranging in age from 55 to 81 years (mean age, 69.7 years). Selective angiography of bilateral external carotid arteries, internal carotid arteries, and vertebral arteries were reviewed, and special attention was paid to the presence of drainage routes through the SPS from CSDAVFs. We also evaluated the normal cerebral venous drainage via the SPS of the affected side in the patients with CSDAVFs.

Normal Hemodynamic Features of the SPS
In 119 patients, the SPSs were partially or totally demonstrated on either carotid angiography or vertebral angiography in 235 of 238 sides. The SPS was not found in 3 sides (Fig 1) on CAG or VAG. The hemodynamic status of the SPS in the patients with normal cerebral venous circulation is summarized in Tables 1 and 2. On CAG, the SPS was demonstrated in 11 sides (4.6%). In 8 sides, the SPS was partially demonstrated to receive blood from the basal vein of Rosenthal via the anastomosis of the lateral mesencephalic vein and the petrosal vein, which included in 6 sides the anterior segment of the SPS with venous drainage from the petrosal vein to the cavernous sinus and in 2 sides the posterior segment with venous drainage from the petrosal vein to the transverse sinus (Fig 2A-D). In 2 sides, the superficial middle cerebral vein was found to terminate into the anterior segment of sphenopetrosal sinus (Fig 2E, -F). In the remaining side, the whole SPS was demonstrated with the blood flow from the transverse sinus into the cavernous sinus via the SPS (Fig 3). On VAG, the SPS was demonstrated to receive blood from the petrosal vein in 235 sides (98.7%). Among these, the whole SPS was demonstrated in 116 sides (48.7%) and partially demonstrated in 119 sides (50%); of the 119 partial demonstrations, it appeared at the posterior segment in 82 sides (34.5%) and at the anterior segment in 37 sides (15.5%) (Figs 4 and 5). Additional findings included variations of the SPS, and duplication of the SPS and disconnection between the anterior segment (5 sides [2%]) and the posterior segments (6 sides [2.5%]) of the SPS was observed (Fig 6). Venous drainage from the cavernous sinus on both CAG and VAG was not demonstrated.

Patients with CSDAVFs
SPS drainage was found in 7 (28%) of 25 patients with CSDAVFs. Table 3 summarizes the hemodynamic status of the SPS in the 7 patients with SPS drainage from the CSDAVFs. Among these 7patients, the CSDAVFs drained through the whole SPS into the transverse sinus in 3 patients (12%) (Fig 7) and through the anterior segment of the SPS into the tributes of the petrosal vein without draining   into the posterior segment of the SPS in 3 patients (12%) (Fig 8). In the remaining patient, the CSDAVF drained through the anterior segment of SPS into the superficial middle cerebral vein (sphenopetrosal sinus), in which duplication of the SPS was also found (Fig 9).
In 18 CSDAVFs without SPS drainage, the SPS was partially observed at the posterior segment alone in 13 patients and at the anterior segment alone in 4 patients. The SPS was not found in 1 patient in whom other cerebellar veins were well developed.

Embryologic Development and Hemodynamic Features of SPS
In the development of the cerebral venous system, the pro-otic sinus participates in forming the cavernous sinus and foramen ovale venous plexus (pterygoid plexus) and receives blood flow from the superior ophthalmic vein. After developing to the    ). Later, the dorsal remnant of the primitive tentorial sinus anastomoses between the cavernous sinus and the SPS at the entrance of the petrosal vein, which then becomes the anterior segment of the SPS. This anastomosis is sometimes plexiform or poorly developed. According to the explanation of the development of the SPS by Padget, 7 the SPS develops primarily as a drainage route from the petrosal vein, receiving blood from the anterior cerebellar and brain stem venous systems. This corresponds well to our results regarding the hemodynamic feature of the SPS. In our results, demonstration of the SPS on carotid angiography is highly associated with the variation of cerebral venous drainage, including hypogenesis of the basal vein of Rosenthal and the sphenopetrosal type of the superficial middle cerebral vein (persistent primitive tentorial sinus), which also supports Padget's explanation. On VAG, approximately 50% of the SPS demonstrated partially, at either the anterior segment or the posterior segment, to receive blood from the petrosal vein. Disconnection of the anterior and posterior segment of the SPS was also seen in a few patients. These findings would represent hypoplasty or malfusion of the anterior and posterior segments of the SPS in development. Duplication of the SPS would be due to the plexiform feature of the remnant of the primitive tentorial sinus.

Relationship of the Types of SPS with SPS Drainage of CSDAVFs
In our results, SPS drainage from CSDAVFs was found in 7 patients (28% of all 25 patients with CSDAVF and 37% of 16 patients with inferior petrosal sinus occlusion). Among the 7 patients with SPS drainage, CSDAVFs drained through the whole SPS from the cavernous sinus in 3 patients. In the other 3 patients, DAVFs drained through the anterior segment of the SPS into the tributes of the petrosal vein without draining into the posterior segment of the SPS. Although stenosis of the dural sinus may develop secondary to high-flow angiopathy caused by the arteriovenous fistulas, SPS drainage and its drainage patterns would be associated with variation of anatomy of the SPS. When CSDAVFs occur in patients with hypoplasty of the posterior segment of the SPS, the fistulas have a higher risk of venous reflux into the cerebellopontine veins compared with the case in which the whole SPS is present. Regarding transvenous embolization, an approach via the SPS is applied in some patients. 8 These variations of segmental hypoplasty and plexiform features of the SPS could affect accessibility via the SPS. Furthermore, multiplication of the SPS may have a risk of residual SPS drainage after transvenous embolization.

CONCLUSIONS
Our study suggests that the SPS normally works as a drainage route receiving blood from the anterior cerebellar and brain stem venous systems and does not work as a normal drainage route from the cavernous sinus. More than half of the SPSs have segmentally hypoplastic features on angiography. The variation of the hemodynamic features would be related to the relatively lower frequency and to the 2 different types of SPS drainage from CSDAVFs.